Computational experiment-aided prescriptive decision-making for complex supply chains: A case of multi-generation smartphone marketing

A supply chain is a complex multivariate evolutionary system that challenges traditional decision-making par-adigms and calls for new decision-making frameworks which can comprehensively support multivariate potential emergence prediction and further extend the decision-making from prediction to prescription. To fill these methodological gaps, a computational experiment-aided prescriptive decision-making framework was proposed for the analysis of complex supply chain evolution. The proposed framework enables the exploration of multi-variate emergences for solution optimization in a closed-loop interactive iteration between computational ex-periments and evolutionary multi-objective algorithms. This framework enhances the solution stability behind multivariate emergences and can support Pareto-optimal solution selection based on a multi-criteria decision -analysis approach. A case study of multi-generation smartphone marketing was then conducted to validate the proposed framework and illustrate its applicability. Results indicate that the proposed framework can (i) bridge the complex nonfunctional relationships between impact factors and objectives for marketing prediction; and (ii) exploit the multivariate marketing performances within these relationships to make optimal prescriptions after the closed-loop interactive iterations. This framework enables managers to obtain Pareto-optimal solution pre-scriptions with improved solution stability surpassing those of traditional prediction.